NBER WORKING PAPER SERIES

URBAN ECONOMICS AND ENTREPRENEURSHIP

Edward L. GlaeserStuart S. RosenthalWilliam C. Strange

Working Paper 15536http://www.nber.org/papers/w15536

NATIONAL BUREAU OF ECONOMIC RESEARCH1050 Massachusetts Avenue

Cambridge, MA 02138November 2009

The Kauffman Foundation supported this paper and all of the essays in this issue. In addition, weare grateful to the Marcel Desautels Centre for Integrative Thinking and the Social Sciences and HumanitiesResearch Council of Canada for research support. We also thank Kristina Tobio and Jiashuo Fengfor research assistance. The views expressed herein are those of the author(s) and do not necessarilyreflect the views of the National Bureau of Economic Research.

© 2009 by Edward L. Glaeser, Stuart S. Rosenthal, and William C. Strange. All rights reserved. Shortsections of text, not to exceed two paragraphs, may be quoted without explicit permission providedthat full credit, including © notice, is given to the source.

Urban Economics and EntrepreneurshipEdward L. Glaeser, Stuart S. Rosenthal, and William C. StrangeNBER Working Paper No. 15536November 2009JEL No. J23,L26,M13,O31,R30

ABSTRACT

Research on entrepreneurship often examines the local dimensions of new business formation. Thelocal environment influences the choices of entrepreneurs; entrepreneurial success influences the localeconomy. Yet modern urban economics has paid relatively little attention to entrepreneurs. This essayintroduces a special issue of Journal of Urban Economics dedicated to the geography of entrepreneurship.The paper frames the core questions facing researchers interested in assessing the local causes andconsequences of entrepreneurship, perturbs a core urban model to incorporate entrepreneurship, andconcludes by offering an agenda for future work on the spatial aspects of entrepreneurship.

Edward L. GlaeserDepartment of Economics315A Littauer CenterHarvard UniversityCambridge, MA 02138and NBEReglaeser@harvard.edu

Stuart S. RosenthalSyracuse UniversityDepartment of Economics426 Eggers HallSyracuse, NY 13244-1020ssrosent@maxwell.syr.edu

William C. StrangeRotman School of ManagementUniversity of Toronto105 St. George St.Toronto, ON M5S3E6CANADAwstrange@rotman.utoronto.ca

I. Introduction

Can the economic history of Detroit be told without Henry Ford and Alfred

Sloan? Would Ford have achieved the same success if he had worked in Houston?

Would Silicon Valley have experienced its remarkable growth without Frederick Terman

and William Shockley? Entrepreneurs often seem to have been significantly influenced

by features of their local economies, and they have often influenced the fates of those

economies. Yet, urban economists have only infrequently looked directly at the local

causes and consequences of entrepreneurship.

Urban economists have not been alone in paying little attention to entrepreneurs.

This is a common feature of economic research after World War II. The general

equilibrium models that came to dominate economics had little room for the

idiosyncrasies of the individuals who started firms. The primary role that Schumpeter

(1934) had assigned to entrepreneurs was largely ignored by mainstream economic

theory. Empirical economists focused more on aggregates and on patterns that held

throughout the economy. But over the past decade, entrepreneurship has become an

increasingly established field that has tried to understand business innovators. This

special issue of the Journal of Urban Economics brings together papers that specifically

focus on the local dimensions of entrepreneurship.

While there has been relatively little formal work on cities and entrepreneurship,

the papers in this volume do not come out of a vacuum. Some urban economists,

notably Vernon and Chinitz, wrote directly about entrepreneurship (i.e., Vernon, 1960;

Chinitz, 1961). Moreover, urbanists from outside of economics, like Jacobs (1969) and

Saxenian (1994), have had important insights about the local roots of entrepreneurship.

In addition, some of the canonical work in urban economics can be interpreted as having

an entrepreneurial dimension. For example, in this paper, we will discuss the

entrepreneurial aspects of urban economic research on agglomeration in general and on

New Economic Geography (NEG) in particular.1

1 Research in agglomeration is surveyed by Duranton and Puga (2004), Rosenthal and Strange (2004), while NEG is surveyed by Fujita et al (1999) and Ottaviano and Thisse (2004).

1

Section II of this essay reviews the role of entrepreneurship in urban economics.

Since entrepreneurship has many dimensions, we begin by discussing the definition of

entrepreneurship. We then turn to early urbanists, who promoted a number of ideas about

entrepreneurship that can help systematize research in this area. Perhaps the most

important idea is that entrepreneurship is important for urban success. Smith (1776) and

Marshall (1920) both seemed to share this view, and it is the central lesson of the work of

Chinitz and Vernon. While this idea may now be a consensus opinion, there is still

surprisingly little statistical work that bears it out.

The other ideas relate to the reasons that different places spawn different levels of

entrepreneurship. It is not at all surprising that economists have linked the level of

entrepreneurship to the returns to entrepreneurship: the supply of entrepreneurship slopes

up. While this idea has been around since Smith, there has been little work actually

measuring the response of entrepreneurship to these financial incentives. One version of

this idea, properly credited to Smith himself, is that large cities have more demand for

specialized products which makes them particularly attractive places for start-ups

creating new products.2 This naturally moves the equilibrium up the entrepreneurial

supply curve.

It is also not surprising that economists have looked for shifters of the

entrepreneurial supply function. There are many local characteristics that might be

responsible for such shifts. Chinitz, Marshall and others emphasized that the level of

entrepreneurship is related to the supply of inputs needed by entrepreneurs, including

material inputs, skilled labor and financing. Chinitz and Marshall both also emphasized

the spread of knowledge as forces that could encourage entrepreneurship. And the list of

entrepreneurial supply shifters should also include political and cultural forces, as well as

natural advantage.

In Section III, we present a brief urban model that incorporates entrepreneurship.

We work with the production and consumption assumptions of Krugman (1991), a model

that shares important features with Abdel-Rahman (1988), Fujita (1988), Rivera-Batiz

(1988). We share with Rivera-Batiz (1988) the focus on a single small city in a large

2See also Duranton and Puga (2001), Duranton and Jayet (2009), and Waldfogel (2009) for related work.

2

open economy. The key formal results in this section document both the impact that

entrepreneurs on local success and also the local factors that influence entrepreneurship.

Section IV then uses this framework of the causes and consequences of

entrepreneurship to categorize the papers in this issue and relate them to other recent

research on entrepreneurship. Many of the papers here pay particularly close attention to

the supply of entrepreneurial people across space. Fewer touch on the effects of

entrepreneurship or the link between entrepreneurship and the returns to entrepreneurial

activity. Section V concludes and sets out the major questions for future research.

II. Entrepreneurship in urban economics

What is an entrepreneur? Webster’s Dictionary (1970, p. 467) defines an

entrepreneur as, “A person who organizes and manages a business undertaking, assuming

the risk for the sake of profit.”3

There are several distinct economic aspects of entrepreneurship. One is that that

entrepreneurs are their own bosses. In this conception of entrepreneurship, entrepreneurs

are self-employed. However, this approach misses the proprietary aspect of

entrepreneurship; entrepreneurs can also be thought of as owners, willing to assume risk

in exchange for returns (Cantillon, 1931; Knight, 1921). At least initially, the firms that

entrepreneurs create are small, so another dimension of entrepreneurship is its role in

small business. But entrepreneurship is not simply about a choice of occupation or about

ownership, it is also fundamentally dynamic. Thus, entrepreneurs can also be conceived

of as being the creators of new firms, and the study of entrepreneurship is the study of

entry. In this view, entrepreneurs are agents of change. That view leads to yet another

possible conception of entrepreneurs as innovators, agents of transformative change

(Schumpeter, 1939), and not simply entrants in a market that is fundamentally the same

year after year. There are, thus, five facets of entrepreneurship: self-employment, small

firms, ownership, entry, and innovation.4

3 Classical economists often used the word “undertaker” to mean essentially the same thing. The alternative meaning of that word has caused it to disappear from use in the twentieth century. 4 All of these aspects of entrepreneurship involve the entrepreneur exercising judgment, a point made by Casson (1982).

3

Since all of these facets of entrepreneurship are important, it would be a mistake

to take the overly narrow view that an entrepreneur must have all of the five

characteristics listed above. Examining the various aspects of entrepreneurship is

logically sensible, and the overly narrow definition excludes important activities. In the

non-profit sector, for instance, entrepreneurs often cede ownership over the enterprise to

an external board. In the for profit sector, shares are typically sold to outsiders, which, in

principle, could lead to a separation of ownership and control. Entrepreneurs can also in

some cases work for someone else. For example, Michael Porter is both an employee of

Harvard University and a management consulting entrepreneur, as well as being a scholar

of entrepreneurship.

Measuring these aspects of entrepreneurship is often difficult. In a sense, every

self-employed person is something of an entrepreneur, but using the self-employment rate

to capture the level of entrepreneurship does no weighting for the size of enterprises, or

the level of risk and innovation. After all, many of the self-employed own nothing but

their own human capital. Measuring entrepreneurship instead with the number of newly

established firms does better at capturing size. At least initially, the firms that

entrepreneurs create are small, so in some cases, entrepreneurship is empirically linked to

an abundance of small firms. However, some conceptions of entrepreneurs conceive of

the entrepreneur as being more than just another business owner. In this view, true

entrepreneurs do more than just open another hot dog stand, they actually do something

new. All of this means that it is difficult for the researcher to capture all of the

potentially relevant aspects of entrepreneurship.

The history of the Journal of Urban Economics allows us to look quantitatively at

the history of the entrepreneurship research in urban economics. An electronic search

brings up the word “entrepreneur” in 57 distinct articles in the Journal of Urban

Economics. 21 of these articles were prior to 1990. In many cases, the word

entrepreneur occurs only once, often in the citation list. The pre-1990 articles that

discuss entrepreneurs at any length generally focus on entrepreneurs who either build

housing or create entire cities, as in Henderson (1985). In the 1970s and 1980s, Journal

of Urban Economics published only two articles that focused primarily on private sector,

non-housing related entrepreneurship (Bates, 1978, Sveikauskas, 1979). The situation is

4

similar at Regional Science and Urban Economics, with 67 papers mentioning

entrepreneurs, 15 before 1990.5

The lack of attention to entrepreneurs is not merely a reflection of editorial

decisions of these two major urban field journals. The pioneering urban economists of

the 1960s, such as William Alonso, Richard Muth, Edwin Mills, and John Kain, rarely

addressed entrepreneurship outside of the housing sector.6 The absence of

entrepreneurs in urban economic papers before 1990 reflects both forces that operated

throughout economics and factors specific to the field. Urban economics arose as a field

in part in response to the raging debates about American urban policy that took place in

the 1960s and 1970s. The field, therefore, tended to focus on problems of housing

markets and urban public policy and not so much on entrepreneurship.

Moreover, the focus of many urban economists has been on creating formal

economic models of cities, following either in the linear programming tradition of

Beckman and Koopmans (1958) or the more continuous tradition of mainstream general

equilibrium models. In these models, entrepreneurs may be embedded in firms, but in the

embedding, much of what was interesting about entrepreneurship disappeared. The

lumpy, random nature of entrepreneurship fit poorly into an agenda aimed at creating

tractable models. Instead, assuming free entry of firms enabled modelers to work with

the powerful zero profit condition that could deliver powerful results. The mathematical

advantages of non-entrepreneurial economics were not unique to the urban field:

entrepreneurs are rarely encountered in models elsewhere in economics.7

For this reason, some of the most important early insights on entrepreneurship in

cities were written by urban economists who used prose rather than algebra. Perhaps the

most important set of insights were generated by the New York Metropolitan Region

Project of the 1950s that brought together, among others, Hoover, Vernon and Chinitz.

These authors’ work (i.e., Hoover and Vernon, 1962; Vernon, 1960; Chinitz, 1961) was

stylistically closer to modern business history, which has consistently focused on

entrepreneurship and place (e.g. Saxenian, 1994), than to formal urban economics. The 5 The Science Direct files on the Journal of Urban Economics go back to 1974 and the journal’s founding; the files on Regional Science and Urban Economics begin in 1975 with Volume 5 of that journal. 6 In this statement, we distinguish between Mills’ research published in journals and his textbook, Mills and Hamilton (1997). The latter does address entrepreneurship outside of the housing sector. 7 See Baumol (1968) for a discussion of this issue. A prominent exception is Lucas (1978).

5

incorporation of entrepreneurship into urban economics really started in the 1970s, with

the work of Sveikauskas and Henderson, who produced a long string of articles that relate

to entrepreneurship. Over the past 20 years, entrepreneurs have steadily become more

important in urban economics, as urban economists have increased their focus on city

economies, developed relevant models, and gotten access to data sets that are relevant to

empirical research on entrepreneurship and innovation.

Yet despite the scarcity of papers with formal algebra or econometrics, the big

ideas about entrepreneurship were sketched long before the current surge in

entrepreneurial research. Perhaps the single most important idea that comes out of the

focus on entrepreneurship in cities is the claim that entrepreneurs play a critical role in

making cities economically dynamic. For example, in Smith’s (1776) discussion of the

rise of cities, dynamic burghers play a leading role. He describes the introduction of

“those manufactures which are fit for distant sale,” which can be interpreted as

technologically advanced goods and credits. He also discusses the benefits of “the

violent operation, if one may say so, of the stocks of particular merchants and

undertakers.” Marshall (1920) tied urban growth and entrepreneurship even more tightly,

writing that “localization and the growth of the system of capitalist undertakers were two

parallel movements, due to the same general cause, and each of them promoting the

advance of the other.” Localization enabled the creation of large quantities of

specialized, tradable products, but that production required “capitalist undertakers,” also

known as entrepreneurs.

The work that grew out of the New York Metropolitan Region Project also

emphasized the importance of entrepreneurship for the success of New York. Chinitz

(1961) compared New York and Pittsburgh, and argued that New York’s greater success

was linked to a more abundant “supply schedule of entrepreneurship.” Likewise,

Hoover and Vernon (1962) depict New York City as an incubator of new business

activity, and connects the continuing strength of that city to its constant flow of new

businesses. Later business scholars such as Porter (1990) and Saxenian (1994) would

echo this message that local success depends on innovative entrepreneurs.

More recently, the connection between entrepreneurship and urban success has

been embraced by a number of urban economists. For example, Duranton and Puga

6

(2001) use the term “nursery cities” to describe places that specialize in creating new

firms and succeed through this innovation. In a related vein, Helsley and Strange (2009)

adapt Lazear's (2005) model of balanced skills to establish that size is not everything. In

order to cultivate entrepreneurship, the entrepreneur's skills and the city's resources must

be in a sense complementary. Another theoretical approach to entrepreneurship is taken

by Forslid and Ottaviano (2002), who add "footloose" entrepreneurs to a New Economic

Geography model. These agents are in fixed supply, and their human capital is required

to form firms. The focus here is not on entrepreneurs per se, but instead on how

including this sort of mobile and scarce human capital in the NEG model can improve

tractability. On the empirical side, Glaeser et al. (1992) documents the connection

between small firms and urban success, and interpret these results as reflecting the

benefits of competition and entrepreneurship. Miracky (1992), Rosenthal and Strange

(2003, 2009) and Glaeser and Kerr (2009) provide similar research along these lines.

While it would be hard to imagine a world in which an abundance of entrepreneurs did

not strengthen the local economy, the literature documenting this effect is still in its

infancy.

Establishing a causal connection between entrepreneurship and local success is

difficult because entrepreneurship, in contrast to the proximity to coal or a good harbor, is

unlikely to an exogenous local variable determined by nature. Those urban economists

who have focused on entrepreneurship, Chinitz and Vernon and Hoover, have seen

entrepreneurship as the reflection of other, deeper forces. Broadly, urban economists

have offered several hypotheses about why entrepreneurship differs across space: (1)

differential returns to entrepreneurship (movement along an entrepreneurial supply

curve), (2) differential availability of inputs to entrepreneurship, including

entrepreneurial human capital and (3) differential supplies of ideas; and (4) differences in

the local culture, political system, or endowments. The first hypothesis refers to a

movement along an entrepreneurial supply curve. Hypotheses (2)-(4) refer to a shift in

an entrepreneurial supply curve. It is worth remarking that this list is parallel to the well-

known list of explanations for the agglomeration of economic activity, a parallel to which

we will return later.

7

Regarding the movement along an entrepreneurial supply curve, hypothesis (1)

above, Smith’s famous dictum that the “division of labor is limited by the extent of the

market,” suggested that certain entrepreneurial activities could only make financial sense

in large communities or places that had ready access to transportation. Smith’s

description foreshadows the modern description of agglomeration economies, where a

larger home market can increase the returns to introducing new products, as in Krugman

(1991). Chinitz, as well, argued that New York’s scale had some role in its

entrepreneurial nature, although he argued that scale alone was not enough to distinguish

Gotham from Pittsburgh.

Nearly fifty years ago, Chinitz (1961) argued that economists ignored

entrepreneurship because “the implicit assumption, I suppose, is that the supply schedule

of entrepreneurship is identical at all locations.” The rest of the ideas in the list relate to

factors that would generate different supply curves for entrepreneurship across locations.

The second theory of local entrepreneurship emphasizes the importance of inputs into

entrepreneurship: some places have more venture capital or the right type of labor or

independent input suppliers. For example, if the bulk of the firms in a region are

vertically integrated then a lack of independent suppliers may make it difficult for a new

firm to sprout. In some industries, skilled labor is vital and as a result a virtuous circle

can occur where entrepreneurs come to a place because of the workers and the workers

come to the place because of the entrepreneurs. Idiosyncrasies in firm outcome may

enhance the gains for entrepreneurs to locate in large cities, since statistical returns to

scale essentially help protect workers and lenders (Helsley and Strange, 1990). Workers

may be more willing to take on the risks of working for an entrepreneurial start-up in a

large city with plenty of alternative employers.

The third hypothesis is that places may differ in the generation and transmission

of entrepreneurial ideas. Marshall (1920) emphasized the role of ideas in infrastructure

and argued that the flow of ideas from person to person was an external economy that

enhanced innovation in cities: “if one man starts a new idea, it is taken up by others and

combined with suggestions of their own; and thus it becomes the source of further new

ideas.” He described an entrepreneurial chain where “subsidiary trades grow up in the

neighborhood, supplying it with implements and materials, organizing its traffic, and in

8

many ways conducing to the economy of its material.” One implication of Marshall’s

logic is that entrepreneurs will congregate next to one another to learn from each other.

Another implication is that we should expect to see chains of innovation where one big

idea is followed with many others.

Jacobs (1969) is strongly associated with the view that new ideas are the well-

spring of entrepreneurship and new businesses make cities grow. Like Chinitz, she

admired small firms, which she saw as being conducive to innovation, and density, which

helps to speed the flow of ideas. She also emphasized industrial diversity, arguing that

some of the biggest innovations are the product of cross-industry fertilization. A modern

example of this phenomenon is Michael Bloomberg, who used his financial expertise

gained at Salomon Brothers to create an information technology firm that could cater to

traders.

It is important to recognize that the inputs, skills, and ideas explanations for the

variation in entrepreneurship can arise endogenously. Regarding skills, Chinitz

suggested that the children of corporate managers might be less likely to become

entrepreneurs than the children of small business owners. In this way, historical

industrial specialization in small firm industries, like the garment trade, might lead to an

ongoing abundance of entrepreneurs, who then encourage their own spinoffs. Similar

stories of positive feedback can be told of inputs and of ideas.

The fourth explanation is that local differences in political system, culture, or in

other local endowments have the potential to impact entrepreneurship. Smith notes that,

“Order and good government, and along with them the liberty and security of individuals,

were ...established in cities at a time when the occupiers of land in the country were

exposed to every sort of violence.” He argues that this security of property led city

residents to take on more business risks. Saxenian (1994) has emphasized local

difference in culture. Silicon Valley investors, for instance, did not blackball

entrepreneurs who had failed previously, a forgiving attitude that is credited with the

Valley's entrepreneurial culture. Finally, natural advantage may impact

entrepreneurship. Pittsburgh’s coal mines made it a center for steel which is inherently

less entrepreneurial while New York’s port gave it access to the world and attracted an

abundance of entrepreneurs

9

As noted above, explanations for differences in entrepreneurial activity parallel

common explanations for agglomeration. This parallel leads naturally to the conclusion

that entrepreneurship can be part of a virtuous circle, where entrepreneurial activity leads

to the circumstances that foster further activity. Of course, the flip side of this conclusion

is that the absence of entrepreneurship can lead to a vicious circle. This strongly suggests

that an improved understanding of entrepreneurship has the potential to help in the

understanding of poverty and urban decline.

II. Entrepreneurship in an urban model

This section will present a simple NEG model that incorporates entrepreneurs.

The model builds on Abdel-Rahman (1988), Fujita (1988), Rivera-Batiz (1988), and

Krugman (1991). None of these paper contains the word entrepreneur, but the NEG

analysis that they offer can be extended in a way that captures the connection between

entrepreneurship and local success and some of the causes of entrepreneurship.

This extension is meant to accomplish three things. First, the analysis will help us

to better understand the equilibrium forces that can explain the spatial variation of

entrepreneurial activity and the impact of entrepreneurship on the larger economy.

Second, the analysis will provide context for the rest of the papers in this issue. Third, by

showing how entrepreneurs can be included in one important urban model, this section

suggests a program for introducing entrepreneurs into other lines of urban research,

pointing towards an agenda for future spatial research on entrepreneurs.

Exogenous entrepreneurs

Following Krugman, individual utility is defined over an aggregate of separate

manufactured goods, each denoted and land, denoted L. Denoting the elasticity of

substitution among the varieties of manufactured goods by σ and adopting Cobb-Douglas

utility, we have

U = . (1)

10

We will focus on one small city within a large open economy. The city has a fixed total

supply of land denoted . The renters live in the city and spend their income there.

The basic technological assumption is that to produce X units of a manufactured

good, a producer requires units of labor. We assume that some manufactured

good varieties are traded, while others are not. We suppose that the price of traded

goods is normalized to (which implicitly assumes that the price of labor external to

this city is normalized to one). The price of non-traded goods will be endogenous.

Krugman critically assumes that there is free entry of firms to the point where

profits are zero. This is, implicitly, an assumption about entrepreneurship: the supply

schedule of entrepreneurs is everywhere the same, and indeed everywhere horizontal.

This assumption essentially takes entrepreneurship out of the model. Instead, we assume

that there are a fixed number of entrepreneurs, denoted . These entrepreneurs produce

two sorts of goods. We suppose that a share of the entrepreneurs produce traded

goods. The external demand for each traded good produced in the city is assumed to be

given by

, (2)

where P is the traded good's price. Let denote the number of entrepreneurs in the

non-traded sector. The local demand for goods is determined endogenously. These

assumptions allow us to look at the city in isolation.

Using this setup, we are able to characterize how entrepreneurship influences

market outcomes for a closed city:

Proposition 1 (Exogenous entrepreneurs in a closed city): If the population of the city is

fixed:

(A) An increase in the number of entrepreneurs will increase wages and worker

utility, decrease the number of workers per firm, and will have an ambiguous

effect on land values.

(B) An increase in the share of entrepreneurs who produce traded goods will

increase wages, worker utility, and land prices, and have no impact on the number

of workers per firm.

11

Proof: See Appendix.

Most of the results in Proposition 1 are intuitive. A larger number of

entrepreneurs increase wages and the welfare of workers. An increase in the share of

entrepreneurs who sell on the global market has similar effects. Places with plenty of

entrepreneurs have a smaller average firm size. The ambiguous connection between the

number of entrepreneurs and land values is somewhat surprising. It reflects the fact that

an increase in the number of entrepreneurs can cause the profitability of firms to decline

if the fixed costs involved in each entrepreneurial activity are sufficiently high.

The primary limitation of Proposition 1 is that it takes the population of the city as

fixed. The hallmark of urban economics is that population levels respond to changes in

parameters. We now, therefore, adopt a specification where, in spatial equilibrium, the

welfare of manufacturing workers in the city must equal a reservation utility level, and

population adjusts accordingly. If the other parameters in this model are fixed, then

worker utility declines with the population level because of competition for land.

In this case, it follows that:

Proposition 2 (Exogenous entrepreneurs in an open city): In an open city:

(A) An increase in the number of entrepreneurs will increase city population and

land values and have an ambiguous effect on wages, the number of workers per

firm, and profits of non-traded good entrepreneurs.

(B) An increase in the share of entrepreneurs who produce traded goods will

increase wages, city population, land prices and the number of workers per firm

and will have an ambiguous effect on profits of non-traded good entrepreneurs.

Proof: See Appendix.

In this case, more entrepreneurship and more traded good entrepreneurship make

the city larger and land in the city more expensive. Entrepreneurship may not, however,

increase wages, since entrepreneurs are making the city more attractive by creating a

12

wider range of goods that can be bought there.8 While it is always true that an increase in

the share of entrepreneurship in traded goods increases firm size, the impact of

entrepreneurship on firm size in this case is ambiguous. It is possible that more

entrepreneurship will attract so many new workers that the number of workers per firm

will actually increase.

The most interesting results connect the level of entrepreneurship with the return

to entrepreneurship. Proposition 2 illustrates that increases in the level of local

entrepreneurship have an ambiguous effect the returns to entrepreneurship. This suggests

the possibility for a virtuous circle where the presence of entrepreneurs increases the

returns for others to be entrepreneurial. The economics behind this ambiguity are

straightforward: more entrepreneurs mean higher wages, which increases costs, but more

entrepreneurship also means a larger market which increases the benefits to

entrepreneurship. Limited land ensures that wages will rise with entrepreneurship, which

helps drive wages up and pushes the returns to entrepreneurship down. If is

sufficiently high, then the land sector is less important, and entrepreneurship is more

likely to lead to a virtuous circle where more entrepreneurs increase the returns to

entrepreneurship.

The main limitation of Proposition 2 -- which applies as well to Proposition 1 -- is

that we have treated the supply of entrepreneurs as being exogenous. Given the well-

documented variation in entrepreneurial activity across locations, we now relax this

assumption and extend the model to consider an endogenous supply of entrepreneurs. So

far we have asked about the implications of entrepreneurship for urban success. We now

turn to the causes of entrepreneurship. The model will show that a range of

circumstances can potentially be related to local entrepreneurial activity. These include:

density, skills, a tradition of entrepreneurship, human capital, and physical capital

Endogenous entrepreneurs

So far, we have treated entrepreneurs as a fixed factor of production that

determines the success of the city. Yet a serious treatment of entrepreneurship in urban

8 Entrepreneurs are acting as amenities in this setup. For recent work on amenities, see Rappaport (2007), Berger, Blomquist, and Peter (2008), Chen and Rosenthal (2008).

13

locations must endogenize the supply of entrepreneurs. The simplest way of

endogenizing entrepreneurship is to assume that a city has an upward-sloping supply

curve of entrepreneurs, defined by a distribution function so that the share of people

whose costs of entrepreneurship are less than X is denoted F(X). As such, the number of

entrepreneurs is defined as F(π), where π denotes the expected profits from becoming an

entrepreneur.9 To simplify matters, we assume that entrepreneurs do not know whether

their product will serve only a local market or will become a traded good when they are

choosing whether or not to become active as entrepreneurs,. As such expected profits

equals times the profits made by traded goods entrepreneurs plus 1 times the

profits made by non-traded goods entrepreneurs.10

The previous section raised the possibility that entrepreneurial profits could be

increasing with the number of entrepreneurs. If that is the case, then it is at least possible

that there are multiple equilibria in entrepreneurship. In one equilibrium, the number of

entrepreneurs is high and so are the returns to entrepreneurship. In the other equilibrium,

the number of entrepreneurs is low and so are the returns to entrepreneurship. While we

do not wish to rule out the possibility of a global change between these two sorts of

equilibrium, in this section we will consider local comparative statics on the number of

entrepreneurs taken a und a stable equilibria, which is defined so that ro

. (3)

Equation (3) essentially means that an increase in entrepreneurship will not have a

destabilizing increase in the profits for each entrepreneur.

With this assumption, the following proposition holds:

Proposition 3 (Endogenous entrepreneurship):

9 This approach has obvious similarities to Lucas (1978). It is worth pointing out that it implicitly assumes that entrepreneurs are risk neutral. In contrast, Kihlstrom and Laffont (1979) build a supply curve of entrepreneurs from differential risk tolerances of economic agents. 10 An alternate approach to endogenizing entrepreneurial activity in a given city is to adopt the Forslid-Ottaviano (2002) "footloose entrepreneurs" model. In our approach, we have assumed immobile entrepreneurs, an assumption that seems consistent with empirical work on entrepreneurship by Sorenson and Audia (2000), Klepper (2007), Dahl and Sorenson (this issue) and others.

14

(A) In a closed city, the level of entrepreneurship is increasing with city

population and decreasing with fixed cost (α) and has an ambiguous relationship

with the share of traded goods and demand (Q).

(B) In an open city, the level of entrepreneurship is increasing with demand (Q)

and land area ( ), is decreasing with fixed cost (α), and has an ambiguous

relationship with the share of traded-goods producers.

Proof: See Appendix.

Many of these results are intuitive. Smaller fixed costs of starting a business

always lead the level of entrepreneurship to rise.11 Higher fixed costs both reduce

profits directly and increase the wage. An increase in city population reduces wage and

increases demand, both of which make entrepreneurship more attractive. In the open city

model, an increase in available land, , causes the population to rise which then indirectly

increases the amount of labor.

The least intuitive results concern the parameters and Q. The first of these

parameters makes it more likely that the entrepreneur will produce a high-value traded

good; the second parameter increases the returns to producing such a traded good.

Intuitively, it would seem that either variable should increase the returns to

entrepreneurship and the number of entrepreneurs. That is exactly what happens in the

case of the open city when Q rises. However, for Q in the closed city case, and for the

variable in either case, the results are ambiguous. These variables cause wages to rise

and it is possible that higher values of Q or can actually drive entrepreneurship down

by pushing wages up so much. While this does not seem likely, these results capture

perhaps the flavor of Detroit after World War II, when the success of the automobile

industry may have crowded out other entrepreneurial activities.

How do these results relate to the core theories about the heterogeneity of

entrepreneurship across space? The variable can be interpreted as reflecting the inputs

11 It is worth observing that the population is mobile, but entrepreneurs are not. Changes in the equilibrium amount of entrepreneurship come from the activation of local entrepreneurs. This seems to us to be consistent with the evidence of entrepreneurial stickiness in Michelacci and Silva (2007) and in Klepper (this issue).

15

needed for entrepreneurship. If those inputs are abundant, perhaps because of large

numbers of input suppliers or venture capitalists, then entrepreneurship will be more

common. The variables N and connect entrepreneurship to the size of the market. In

bigger areas, there will be more entrepreneurs, but in this model there is no assurance that

the level of entrepreneurship will rise by more than one-to-one. This could be assured by

assuming a sufficiently elastic supply of entrepreneurs. The variables of Q or reflect

exogenous factors in the place that increase the potential returns to entrepreneurship.

Historically, perhaps, New York’s port might have been seen as causing an increase in

the value of Q.

This framework illustrates the impact on entrepreneurship of market size, returns

to entrepreneurship and input costs. It also illustrates the impact of entrepreneurship on

local economic performance. A virtuous circle arises through market size. One could

similarly add entrepreneurs to models that emphasize inputs, skills, or ideas to obtain a

parallel sort of circularity (see, for instance, Strange et al, 2006).

The framework does not address the Chinitz (1961) intuition that the supply of

entrepreneurs might be different across space. To capture this possibility, we can simply

assume that the number of entrepreneurs equals , where shifts the supply

schedule across space. This would capture the possibility that some places just have

more entrepreneurial people. Alternatively, urban density might act as a multiplier so

that the supply becomes , where 1, because each entrepreneur spreads ideas to

others. Unsurprisingly, then it follows that the level of entrepreneurship is rising with

either or (so long as 1 . While all of our other comparative statics caused

the level of entrepreneurship to increase by increasing the net returns to entrepreneurship,

an increase in these variables will cause the net returns to entrepreneurship to fall, at least

if the conditions needed for the returns to entrepreneurship to be declining in the level of

entrepreneurship hold. However, it is possible, since the returns to entrepreneurship can

actually increase with the number of entrepreneurs, that this increase in supply may

actually be associated with an increase in the returns to entrepreneurship.

IV. The Causes and Consequences of Entrepreneurship

16

We now turn to the papers in this issue and the literatures from which they

emerge. The section will largely be devoted to discussing how both this body of research

fits with the causes and consequences of entrepreneurship framework laid out in this

paper.

The Impact of Entrepreneurship

Both in this issue and more widely, there has been relatively little literature on the

broad impacts on entrepreneurship on urban economies.12 We suspect that this reflects

two problems. First, the vast majority of entrepreneurship researchers, and indeed

probably most of the academic community, rarely question the positive benefits of

entrepreneurship for the local community. After all, real world examples abound of

entrepreneurs who have helped their cities economically. Second, the ability to find a

causal link between entrepreneurship and urban success would require exogenous

variables that increase entrepreneurship but have no other impact on the local economy.

As Section III's model illustrates, most of the candidate variables that explain

entrepreneurship, such as market size, could easily have a direct impact on city growth.

The piece of evidence that is most suggestive of the positive benefits of

entrepreneurship on city growth is the strong connection between small average firm size

and subsequent growth (see Glaeser et al., 1992; Rosenthal and Strange, 2003 and 2009;

and Glaeser et al, this issue). This correlation is certainly suggestive. Small firms are

likely to be newer and more entrepreneurial, but after all, this correlation has many

interpretations. Small average firm sizes also means more competitive labor markets, or

perhaps fewer regulatory barriers to growth.13,14

Audretsch and Keilbach (2004) have used an alternative measure of

entrepreneurship—the rate of new start-ups. This variable is strongly correlated with the

economic output of West German Counties. The start-up rate in the high tech sector is

particularly associated with success. Just as in the case of the firm size results, these

12 There has been a larger body of work on entrepreneurship and economic success at the country level, as in Audretsch and Thurik (2001) and Audretsch et al (2006). 13 The self-employment rate also, far more weakly, predicts urban employment growth (Glaeser, 2007). 14 The result that Wal-Mart has negative impacts on the local economy (Neumark et al, 2008)) can be seen as complementary to the small firm effect. See Haltiwanger et al (this issue) for an analysis of big box firms more generally. See Basker (2005) for evidence that Wal-Mart's effects are not necessarily negative.

17

correlations cannot be conclusively interpreted as causal results showing the benefits of

entrepreneurship, since start-up rates are unlikely to be exogenous.

As discussed above, another aspect of entrepreneurship is the independence of

entrepreneurs. Rosenthal and Strange (2003) consider this directly by distinguishing

between firms that are subsidiaries of another corporate entity from those that are not.

No consistent relationship is found between the presence of non-subsidiary firms and

growth. This preliminary evidence suggests that in thinking about the impacts of

entrepreneurship, the ownership dimension might not be as important as the size

dimension.

Economists have thought, at least since Solow (1956), that technological

innovation is associated with growth. At the local level, Porter (1990) has argued that the

innovativeness of certain clusters accounts for their growth. Moreover, the history of

technological innovations, such as the assembly line and software, suggest a significant

role for entrepreneurs, such as Henry Ford and Bill Gates. Yet there is surprisingly little

formal econometric work at the city level that has quantified the relationship between

local innovation and other outcomes. This is presumably explained by the two points

raised above: no one doubts such a relationship, and identification is difficult to achieve.

Kolko and Neumark (this issue) present an alternative means of assessing the

benefits of different forms of entrepreneurship. Instead of looking at economic growth or

level of output, they focus on the behavior of firms in response to shocks. Many local

leaders would like firms that keep employment steady even during a downturn, and in

principle, local entrepreneurship could provide a cushion against recession. Indeed,

many communities have specifically tried to protect locally owned businesses from

externally owned competitors (such as Big Box retailers) with the idea that locally owned

businesses are more likely to provide stable employment for local workers.

Kolko and Neumark find mixed evidence for this claim. Company headquarters

are more stable than other establishments. Locally owned single establishment firms are

actually more sensitive to downward industry shocks, but less sensitive to downward

regional shocks. Likewise, among smaller firms there is some tendency of local

ownership to mute the response to regional shocks. However, among larger firms, local

ownership has no impact in the response to shocks.

18

The Sources of Entrepreneurship

There has been more research that bears on the sources of entrepreneurship than

on its impact. The line of research in urban economics that has the most to say regarding

the sources of entrepreneurship is the literature on agglomeration. As noted above, this

literature has identified inputs, skills, and ideas as being involved in the process

generating agglomeration economies.15 The literature has also looked at political and

cultural forces and also natural advantage as explanations for agglomeration.

Many themes in this literature relate to entrepreneurship. Fallick et al (2006) and

Freedman (2008) consider the relationship between agglomeration and job hopping.

They show that mobility increases with industrial concentration. This relates to both the

self-employment and change dimensions of entrepreneurship. Holmes (1999) shows a

relationship between vertical integration and agglomeration in the U.S. Li and Lu (2009)

show a similar relationship in China. A large literature has shown a relationship between

agglomeration and innovation. Recent contributions include Agrawal et al (2008),

Gerlach et al (2009), and Simonen and McCann (2008).

In this issue, most of the papers deal with the factors that explain why

entrepreneurship might differ across space. One of these factors is differences in the

returns to entrepreneurship. The only paper in this volume that directly addresses returns

to entrepreneurship is Glaeser et al (this issue). This paper specifically looks at whether

the value of shipments per employee, a proxy for the returns to entrepreneurship, are

higher in places with abundant small firms where there are lots of entrepreneurs. No

correlation is found. This suggests that these clusters of entrepreneurship are not being

created by the presence of unusually high returns to entrepreneurial activity. This clearly

calls for looking at supply shifters.

Chinitz’s (1961) seminal work on entrepreneurship emphasized the importance of

inputs to entrepreneurship, which seemed to be far more prevalent in New York than in

Pittsburgh. He argued that a competitive economy, made up of small independent firms,

would make it much easier for entrepreneurs to find independent suppliers. Certainly, no

input is more important for new entrepreneurs than finance itself, so the availability of

15 See the recent literature reviews in Duranton-Puga (2004) and Rosenthal-Strange (2004).

19

venture capital should surely be one of those variables that could impact the rate of

entrepreneurship.

Chen et al (this issue) look specifically at the geography of the venture capital

industry. They find that this industry is overwhelmingly located in three cities: New

York, San Francisco and Boston. This might seem to give a great advantage to

entrepreneurs working in those locations. However, they also find that these firms

frequently invest outside of their cities and actually earn higher returns from spatially

distant investments. This may, of course, reflect a selection process. If it were more

expensive to invest in distant projects, venture capitalists would tend to invest in only the

most promising of such projects, and this would elevate the return on investments outside

of the firms’ immediate locations, on average. In any case, venture capitalists seem to be

capable of making wise investments away from their home towns, which suggests that

the availability of capital might not be as geographically concentrated as the venture

capitalists.

However, these findings still continue to suggest that the agglomerations of

venture capitalists help create agglomerations of entrepreneurs. If investors only bother

to invest far away if they can get higher returns by doing so, then firms in Silicon Valley

can obtain venture capital for worse projects than firms in Idaho, and the Silicon Valley

firms consequently enjoy an advantage. Somewhat paradoxically, the presence of high

returns elsewhere actually supports the idea that entrepreneurs enjoy an advantage from

being close to the clusters of venture capital.

The paper by Agrawal et al (this issue) takes a different angle in examining the

geography of innovation by looking at the advantage of competition vs. monopoly. Just

as large vertically integrated firms might not sell goods to new start-ups, such large firms

might not provide as many intellectual spillovers if they are more closed to outsiders.

This paper looks at the patenting activity of firms in “company towns” that are dominated

by a single large enterprise and compares that activity with innovation in more

competitive environments.

The main conclusion of Agrawal et al is that the dominant firms in company

towns are, indeed, more inward looking. They are more likely to cite their own patents

than comparable firms elsewhere. Yet these large companies do not seem to reduce the

20

tendency of their smaller neighbors to cite broadly. Moreover, the patents invented in

these company towns have just as much impact as patents invented elsewhere. These

findings could be interpreted as evidence that the creative myopia associated with the

presence of large firms in company towns does not necessarily deter productivity and

innovation.

The paper by Glaeser et al (this issue) also looks at the connection between local

industrial structure and employment growth. They find that an abundance of small firms

strongly predicts new establishment births. As noted above, this relationship does not

appear to stem from an effect of nearby small firms on the returns to entrepreneurship.

Rather, the relationship depends on the supply of entrepreneurs.

A different perspective on the interaction between large and small firms is

provided by Haltiwanger et al (this issue). They examine the degree to which big-box

stores – both retailers and restaurants – displace employment at local smaller, mom and

pop stores. They find that the presence of big box stores adversely affects employment at

smaller firms, primarily by causing smaller firms to exit. Importantly, this effect is

concentrated among small firms that are in the same industry as the big-box competitor.

In addition, this effect attenuates sharply with distance, and is much less pronounced just

a few miles away. The very local geographic nature of the effects of competition from

big box retailers parallels previous research showing that the spillover effects of nearby

agglomerations of employment attenuate rapidly with distance (e.g. Rosenthal and

Strange (2003, 2008), Andersson et al (2009)).

Entrepreneurs are often educated, so an increase in the share of educated workers

can be seen as an increase in the supply of entrepreneurs. An abundance of educated

workers may also increase the returns to entrepreneurship by providing skilled labor, a

necessary input into many firms. The paper by Doms et al (this issue) examines the

connection between the education level of workers themselves and of the workers'

communities and entrepreneurship. It draws on several different surveys, each with its

strengths and weaknesses. This includes a new and unique panel of several thousand

newly established small businesses that are followed for four years (the Kauffman Firm

Survey, or KFS). The paper also utilizes individual-level cross-section data from the

voluminous 5 percent Public Use Micro Sample of the 2000 decennial census.

21

Based on these and other sources, Doms, Lewis and Robb (this issue) first

document a strong positive correlation between the average level of education in a

metropolitan area and the level of education among the area’s community of business

owners and self-employed. This is not surprising, but reinforces the importance of

distinguishing between the influence of individual versus metropolitan-wide levels of

education. Indeed, the paper further shows that while entrepreneurial activity and

business turnover, as measured by new business formation and deaths, is more prevalent

in educated cities, the relationship appears to be most closely associated with individual

rather than city-wide levels of education. Based on the census data, more highly

educated individuals are more likely to be self-employed, and conditional on individual

education, there is no additional positive association between self-employment and

metropolitan area college share.

Are these relationships between education and the propensity for entrepreneurship

mirrored in the returns to entrepreneurship? Doms, Lewis, and Robb (this issue) provide

evidence on this point as well. Specifically, in both the KFS and census surveys, they

find that entrepreneurs with more education enjoy improved business outcomes, and that

this relationship is highly non-linear: there is a strong positive premium for having a

college degree. However, conditional on the individual’s own level of education, the

association between local area education and business outcomes is less clear. In the KFS,

this latter relationship is positive, but not distinguishable from zero. In the census data, a

clear positive relationship emerges: self-employed individual earn more when operating

in more highly educated locations, even after controlling for an extensive array of

industry-metro area fixed effects, the individual’s own level of education, and many other

individual-level controls. Moreover, from both data sources, and especially from the

census, there is suggestive evidence that it is primarily high-skilled sectors that benefit

from the presence of nearby college educated workers. These findings complement

Glaeser et al (this issue) who show that labor-intensive firms are particularly more likely

to form in high human capital areas. They are also related to Bacolod et al (2009) who

find that the returns to skills rise with city size. More generally, these findings confirm

that education plays a crucial but complicated role in contributing to a successful

entrepreneurial environment.

22

Since the location of skilled workers and technical expertise seems to be so

important, it is natural to focus on the mobility of technical workers and innovative

activity. Kerr (this issue) and Dahl and Sorenson (this issue) each shed light here. Kerr

examines the speed with which innovative activity migrates across cities in the United

States. He demonstrates that development of breakthrough technologies tends to be

followed by subsequent intensive research and development that refines the initial

innovation. This process attracts additional scientists and engineers to the location where

the breakthrough occurs, and contributes to movement of innovative activity across cities.

The speed with which such adjustments occur depends in part on the mobility of

technical workers. Kerr demonstrates this by focusing on immigrant scientists and

engineers, a group thought to be particularly mobile and footloose. Evidence confirms

that industries reliant on immigrant technical workers exhibit faster migration of

innovative activity towards locations where breakthroughs occur.

Kerr’s (this issue) work highlights both the dynamics of breakthroughs, as well as

the impact of worker mobility on the speed with which innovative activity and technical

ability spreads across cities. This relates to prior work on the dynamics of industry

migration of across cities. Duranton (2007) and Findeisen and Sudekum (2008) show

that the size distribution of cities in a given country tends not to change much over time,

and that the relative size of individual cities changes only very slowly. However, the

industrial composition of employment within individual cities changes at a comparatively

rapid pace as industries migrate across locations. Kerr’s results highlight factors that

contribute to such migration.

The paper by Dahl and Sorenson (this issue) also considers the mobility of

scientific and technical workers, in this case in Denmark. A key finding is that scientific

and technical workers in Denmark are heavily drawn to locations close to family and

friends. While economic incentives matter – and especially for older workers – Dahl and

Sorenson provide compelling evidence that Danish scientists and technical workers are

willing to trade off substantial income for the opportunity to locate closer to their parents,

high school friends, and other important social contacts. Such ties tend to create

geographic frictions within high-skilled labor markets, and may serve to slow the rate at

which innovative activity would migrate across cities. In many respects, these findings

23

complement the evidence offered by Kerr (this issue) in which mobile workers accelerate

the migration of innovative activity.

The causes of entrepreneurship considered thus far do not come close to forming

an exhaustive list. There are many additional factors, many relating to government

policies, especially regulations that increase the costs of entrepreneurship. Becker and

Henderson (2000) and List et al (2003) both show that air quality regulations reduce firm

birth in the relevant industries. Bertrand and Kramarz (2002) show that French zoning

regulation reduces new firm entry in the retail sector. On the other hand, inaction in the

face of urban problems such as congestion seems to be negatively associated with growth

(Hymel, 2009).

Rosenthal and Ross (this issue) consider a different urban problem, the local rate

of violent crime. Pope (2008) and Linden and Rockoff (2008) show that the fear of crime

can have a large impact on housing prices. Rosenthal and Ross consider the relationship

of crime to entrepreneurship. The paper deals with the sorting of sectors of the economy

into high- and low-crime neighborhoods depending on a sector’s relative sensitivity to

crime. The paper illustrates this by comparing retail industries to wholesale sectors and

high-end restaurants to low-end eateries. Because retail industries rely on pedestrian

shoppers, they will be especially sensitive to violent crime. Because high-end restaurants

do a disproportionate amount of their business in the evening, they should be especially

sensitive to violent crime over the prime dinner hours. Using data for five U.S. cities,

Rosenthal and Ross obtain evidence consistent with these priors: higher levels of violent

crime reduce the share of retail trade in an area relative to wholesale trade, and higher

local rates of violent crime during peak dining hours reduce the presence of high-end

restaurants relative to lower-tier eateries.

Many of the papers in this issue have dealt with the supply of entrepreneurs. Of

course, the local supply of entrepreneurs only matters if entrepreneurs are tied to a local

area. Klepper (this issue) offers a further perspective on this topic by examining the

history of spinoffs in Detroit – with the auto industry – and in Silicon Valley – with the

integrated circuit or semi-conductor industry. Remarkably, the paper draws upon data for

the entire history and lineage of modern day auto makers and producers of integrated

circuits.

24

Klepper documents that early in the auto industry’s history, several locations were

thriving centers of production, as was also the case for integrated circuits. However,

spinoffs from unusually successful founding (parent) companies were themselves more

likely to be unusually successful, and these spinoffs tended to locate close to their parent

firms. Klepper argues that this tendency for overachieving parent companies to spawn

nearby overachieving spinoffs accounts for the eventual dominance of Detroit in the auto

industry, and the corresponding dominance of Silicon Valley for integrated circuits. The

evidence presented by Klepper is compelling and speaks directly to the important role

that unusually talented and innovative entrepreneurs may have on an industry, and the

role of spinoffs in fostering agglomerations of both innovation and industrial activity.

It is less clear why these spinoffs remained in such close proximity to their parent

companies. Was it because of the desire to remain close to family and friends, as in Dahl

and Sorenson (this issue)? Was there or some other aspect of the technical labor force

that might have deterred migration (which would contrast to the role of immigrant

workers as in Kerr, this issue)? Or perhaps it was the traditional role of agglomerative

spillovers in the form of input sharing, labor pooling, and the opportunity to learn from

one’s neighbors, as emphasized in the agglomeration literature. These are issues that

bear further study, but which do not change the important implication of Klepper (this

issue) that any assessment of the impact of entrepreneurship on local urban economies

must take seriously the role of spinoffs and “organizational reproduction.”

V. Conclusion: directions for future research

We believe that the essays in this issue contribute to our understanding of the

major questions about entrepreneurship and economic geography. Yet economic

research is still just beginning to understand the key topics laid out in this paper. Our

ability to model and estimate the creative sparks of Henry Ford and Alfred Sloan remains

limited. In conclusion, we will lay out some of the most important open questions. We

hope that this may guide future research in this area.

First, and perhaps most importantly, what is the impact of entrepreneurship at the

local level? We still lack compelling evidence on the impact of entrepreneurship on

25

cities and regions. The biggest handicap to work in this area is the absence of exogenous

variation in entrepreneurship that is independent of other sources of economic success.

Both basic research and public policy have a great need for definitive work in this area.

Second, what are the causes of spatial variations in entrepreneurial activity?

While there has been much more written about the causes of entrepreneurship than about

its effects, there are still many unanswered questions. To what extent are differences in

entrepreneurship results of differences in entrepreneurial returns? There is evidence that

entrepreneurship increases with the availability of relevant inputs, but we do not know

which inputs are most important. For many specific inputs, we still lack well identified

estimates of the impact of input availability on entrepreneurship. We know more about

how skills and entrepreneurial types impact entrepreneurship. Still, there are many

unanswered questions in this area as well. To what extent do social interactions in a place

create a local multiplier in entrepreneurship? To what extent does events early in an

individual’s career influence the propensity to become an entrepreneur? Finally, how do

government policies and culture impact entrepreneurial activity. Overall, although we

know that the supply curve of entrepreneurs slopes up, we do not understand in a general

way the slope of the entrepreneurial supply curve or how it differs across metropolitan

areas. This remains a crucial area for future research.

Third, whether one considers the local causes of entrepreneurship or the local

effects, it is unclear at what geographic scale the spatial mechanics of entrepreneurship

operate. The papers in this issue have all considered the local dimensions of

entrepreneurship. In most cases, the spatial issues operate at the city level. In this spirit,

several papers consider the city-level conditions that lead to the creation of self-

reinforcing entrepreneurial clusters. In other cases, the spatial issues operate at larger

(regional) or smaller (neighborhood) levels of geography. The research question that is

suggested by these differences in approaches is clear: at what spatial scale do these

entrepreneurs operate? Are the effects highly local, or do they impact entire regions?

Fourth, although the papers in this issue have largely focused on business creation

and economic growth, it is almost certainly true that entrepreneurship has much broader

impacts. Which suggests the fourth question: how does entrepreneurial activity impact

key urban issues? For instance, new business creation has taken place disproportionately

26

27

at the edges of most cities in recent years. This suggests a link between entrepreneurial

patterns and spatial decentralization and sprawl. Similarly, some areas within cities have

robust business sectors. Some areas do not. How do the factors determining

entrepreneurship impact spatial patterns of inequality and ghettoes? Finally, the

competition among local governments discussed by Tiebout (1956) is itself a sort of

entrepreneurship. How does fiscal entrepreneurship impact cities? Such

entrepreneurship has the potential to be particularly important in developing countries

(Lichtenberg and Ding, 2009). There are fine examples of research on all these

questions, but definitive answers have eluded us.

These four large open questions quite naturally suggest a fifth: what are

appropriate policies towards spatial entrepreneurship? Clearly, with only tentative

answers to the first four questions, it is difficult to argue for a comprehensive and

intrusive program. But given the current evidence that entrepreneurship matters and that

there are a range of factors that impact entrepreneurial activity, it would be a grave

mistake to simply ignore entrepreneurship.

Entrepreneurship is rooted in a place, even in the industries that are most

technologically advanced. Few people would doubt that Silicon Valley or New York or

Bangalore have special characteristics that help make them centers for entrepreneurship

in different sectors. Few people would also doubt that entrepreneurs have often played a

major role in forging local economies. For these reasons, the returns are high to bringing

entrepreneurship more squarely into urban economics.

References

Abdel-Rahman, H.M., 1988, Product differentiation, monopolistic competition, and city size, Regional Science and Urban Economics 18, 69-86. Agrawal, A., Cockburn, I., Rosell, C., 2009. Not invented here: Creative myopia and company towns, Journal of Urban Economics, this issue. Agrawal, A., Kapur, D., McHale, J., 2008. How do spatial and social proximity influence knowledge flows? Evidence from patent data, Journal of Urban Economics, 64. Andersson, R., Quigley, J., Wilhelmsson, M., 2009, Urbanization, productivity, and innovation: Evidence from investment in higher education, Journal of Urban Economics, 66, 2-15. Audretsch, K., Keilbach, M., Lehmann, L., 2006. Entrepreneurship and Economic Growth, Oxford University Press. Audretsch, D. B., Keilbach, M. Entrepreneurship capital and economic performance, Regional Studies 38.8 (Nov 2004): 949(11). Audretsch, D., Thurik, R, 2002. Linking entrepreneurship to growth, OECD Science, Technology and Industry Working Papers, 2001/2, OECD Publishing Bacolod, M., Blum, B., Strange, W., 2009. Skills in the city, Journal of Urban Economics 65, 136-153. Basker, E., 2005. Selling a cheaper mousetrap: Wal-Mart's effect on retail prices, Journal of Urban Economics 58, 203-229. Bates, T., 1978. Profitability in traditional and emerging lines of black business enterprise, Journal of Urban Economics 5, 154-171. Baumol, W.J., 1968. Entrepreneurship in economic theory, American Economic Review 58, 64-71. Becker, R., Henderson, V., 2000. Effects of air quality regulations on polluting industries, Journal of Political Economy 108, 379-421. Berger, M., Blomquist, G., Peter, K., 2008. Compensating differentials in emerging labor and housing markets: Estimates of quality of life in Russian cities, Journal of Urban Economics 63, 25-55. Bertrand, M., Kramarz, F., 2002. Does entry regulation hinder job creation? Evidence from the French retail industry, Quarterly Journal of Economics 117, 1369-1413.

28

Cantillon, R., 1755/1931. Essai sur la nature cu commerce en general, ed. H. Higgs, MacMillan, London. Casson, M. C., 1982. The Entrepreneur:An Economic Theory, Edward Elgar, London. Chen, H., Gompers, P., Kovner, A., Lerner, J., 2009. Buy local? The geography of successful and unsuccessful venture capital expansion, Journal of Urban Economics, this issue. Chen, Y., Rosenthal, S., 2008. Local amenities and life-cycle migration: Do people move for jobs or fun?, Journal of Urban Economics 64, 519-537. Chinitz, B.J., 1961. Contrasts in agglomeration: New York and Pittsburgh, American Economic Review 51, 279-89. Dahl, M., Sorenson, O., 2009. The migration of technical workers, Journal of Urban Economics, this issue. Doms, M., Lewis, E., Robb, A., 2009. Local Labor Market Endowments, New Business Characteristics, and Performance, Journal of Urban Economics, this issue. Duranton, Gilles, 2007, Urban evolutions: The fast, the slow, and the still, American Economic Review 91, 197-221. Duranton, G., Jayet, H., 2009. Is the division of labour limited by the extent of the market? Evidence from French cities. Unpublished mimeo. Duranton, G., Puga, D., 2001. Nursery cities: Urban diversity, process innovation, and the life-cycle of products, American Economic Review 91, 1454-1477. Duranton, G. and Puga, D., 2004. Micro-foundations of urban agglomeration economies, in: Henderson, J.V., Thisse J.-F. (Eds.), Handbook of Urban and Regional Economics, Vol. 4, Elsevier, Amsterdam, pp. 2063-2118. Fallick, B., Fleischman, C. and Rebitzer, J.B., 2006. Job hopping in the Silicon Valley: The micro-foundations of an industry cluster, Review of Economics and Statistics 88, 472-481.

Findeisen, S., Sudekum, J. 2008. Industry churning and the evolution of cities: Evidence for Germany, Journal of Urban Economics 64, 326-339. Forslid, R., Ottaviano G.I.P., 2002, An analytically solvable core-periphery model, Journal of Economic Geography 3, 229-240. Freedman, M.L., 2008. Job hopping, earnings dynamics, and industrial agglomeration in the software publishing industry, Journal of Urban Economics 64, 590-600.

29

Fujita, M., 1988. A monopolistic competition model of spatial agglomeration, Regional Science and Urban Economics 18, 87-124. Fujita, M., Krugman, P., and Venables, T., 1999, The Spatial Economy: Cities, Regions, and International Trade, MIT Press, Cambridge. Gerlach H., Ronde, T., Stahl, K., 2009. Labor pooling in R&D intensive industries, Journal of Urban Economics 65, 99-111. Glaeser, E.L, Kallal, H. D., Scheinkman, J. A., Shleifer, A., 1992. Growth in cities, Journal of Political Economy 100, 1126-1152. Glaeser, E.L., 2007. Entrepreneurship and the city. National Bureau of Economic Research Working Paper 13551. Glaeser, E.L. Kerr, W.R., 2009. Local industrial conditions and entrepreneurship: How much of the spatial distribution can we explain?, National Bureau of Economic Research. Working Paper 14407. Glaeser, E., Ponzetto, G., Kerr, W., 2009. Geographic amenities and the agglomeration of innovation entrepreneurship, Journal of Urban Economics, this issue. Haltiwanger, J., Jarmin, R., Krizan, C.J., 2009. Mom & Pop meet big box: Complements or substitutes? Journal of Urban Economics, this issue. Helsley, R. W., Strange, W. C., 1990. Agglomeration economies and matching in a system of cities, Regional Science and Urban Economics 20, 189-212. Helsley, R.W., Strange, W.C., 2009. Entrepreneurs and cities: Complexity, thickness, and balance, Working Paper. Henderson, J. V., 1985. The impact of zoning policies which regulate housing quality, Journal of Urban Economics 18, 302-312. Holmes, T. J., 1999. Localization of industry and vertical disintegration, Review of Economics and Statistics 81, 314-25. Hoover. E., Vernon, R., 1959. Anatomy of a Metropolis, Harvard University Press, Cambridge. Hymel, K., 2009. Does traffic congestion reduce employment growth?, Journal of Urban Economics 65, 127-135. Jacobs, J., 1969. The Economy of Cities, Vintage, New York.

30

Kerr, W., 2009. Breakthrough inventions and migrating clusters of innovation, Journal of Urban Economics, this issue. Kihlstrom, R. E., Laffont, J.-J., 1979. A general equilibrium entrepreneurial theory of firm formation based on risk aversion, Journal of Political Economy 87, 719-48. Klepper, S., 2007, Disagreements, spinoffs, and the evolution of Detroit as the capital of the U.S. automobile industry, Management Science 53, 616-631. Klepper, S., 2009. The original and growth of industry clusters: The making of Silicon Valley and Detroit, Journal of Urban Economics, this issue. Knight, F. H., 1921. Risk, Uncertainty, and Profit, Houghton-Mifflin, Boston. Kolko, J., Neumark, D., 2009. Does local business ownership stabilize employment? Journal of Urban Economics, this issue. Koopmans, T.C., Beckmann, M., 1957. Assignment problems and the location of economic activities, Econometrica 25, 53-76. Krugman, P., 1991. Increasing returns and economic geography, Journal of Political Economy 99, 483-499. Lazear, E.P., 2005. Entrepreneurship. Journal of Labor Economics 23, 649-680. Li, B., Lu, Y., 2009. Geographic concentration and vertical disintegration: Evidence from China, Journal of Urban Economics 65, 294-304. List, J.A., Millimet, D.L., Fredriksson, P.G., McHone, W.W., 2003. Effects of environmental regulations on manufacturing plant births: Evidence from a propensity matching score estimator, Review of Economics and Statistics 85, 944-952. Lichtenberg, E., Ding, C., 2009. Local officials as land developers: Urban spatial expansion in China, Journal of Urban Economics 66, 57-64. Linden, L., Rockoff, J.E., 2008. Estimates of the impact of crime risk on property values from Megan's Laws, American Economic Review 98, 1103-1127. Lucas Jr., R.E., 1978. On the size distribution of business firms, The Bell Journal of Economics 9, 508-523. Marshall, A., 1890/1920. Principles of Economics, MacMillan, London. Michelacci, C., Silva, O., 2007. Why so many local entrepreneurs?, Review of Economics and Statistics 89, 615-633.

31

Millimet, D.L., Fredriksson, P.G., McHone, W.W., 2003. Effects of environmental regulations on manufacturing plant births: evidence from a propensity score matching estimator, Review of Economics and Statistics 85, 944-952. Mills, E.S., Hamilton, B.W., 1997. Urban Economics, 6th ed., Prentice Hall, New York.

Miracky, W.F., 1992. Technological spillovers, the product cycle, and regional growth, Working paper. Neumark, D., Zhang, J., Ciccarella, S., 2008. The effects of Wal-Mart on local labor markets, Journal of Urban Economics 63, 405-430. Ottaviano, G.I.P. and Thisse, J.-F., 2004. Agglomeration and economic geography, in: Henderson, J.V., Thisse J.-F. (Eds.), Handbook of Urban and Regional Economics, Vol. 4, Elsevier, Amsterdam, pp. 2564-2608. Pope, J.,C., 2008. Fear of crime and housing prices: Household reactions to sex offender registries, Journal of Urban Economics 64, 601-614. Porter, M., 1990. The Competitive Advantage of Nations, The Free Press, New York. Rappaport, J, 2007. Moving to nice weather, Regional Science and Urban Economics, 37, 375-398.

Rivera-Batiz, F.L., 1988. Increasing returns, monopolistic competition, and agglomeration economies in consumption and production, Regional Science and Urban Economics 18, 125-153. Rosenthal, S. S., Strange, W.C., 2003. Geography, industrial organization, and agglomeration, Review of Economics and Statistics 85, 377-393.

Rosenthal, S. S., Strange, W.C., 2004. Evidence on the nature and sources of agglomeration economies, in: Henderson, J.V., Thisse J.-F. (Eds.), Handbook of Urban and Regional Economics, Vol. 4, Elsevier, Amsterdam, pp. 2119-2172.

Rosenthal, S. S., Strange, W.C., 2008. The attenuation of human capital spillovers, Journal of Urban Economics 64, 373-389.

Rosenthal, S. S., Strange, W.C., 2009. Small establishments / big effects: Agglomeration, industrial organization, and entrepreneurship, in: E.L. Glaeser, ed., Agglomeration Economics, National Bureau of Economic Research, Cambridge, in press.

Ross, A., Rosenthal, S., 2009. Violent crime, entrepreneurship and vibrant cities, Journal of Urban Economics, this issue. Saxenian, A., 1994. Regional Advantage: Culture and Competition in Silicon Valley and Route 128, Harvard University, Cambridge, MA.

32

Simonen, J., McCann, P. 2008. Firm innovation: The influence of R&D cooperation and the geography of human capital inputs, Journal of Urban Economics 64, 146-154 . Schumpeter, J. A., 1934. The Theory of Economic Development, Harvard University Press, Cambridge, MA. Schumpeter, J. A., 1939, Business Cycles, McGraw-Hill, New York. Smith, A., 1776/2001. An Inquiry into the Nature and Causes of the Wealth of Nations, Bartleby, New York. Solow, R. M., 1956. A contribution to the theory of economic growth, Quarterly Journal of Economics, 70, 65-94. Sorenson, O., Audia, P.G., 2000, The social structure of entrepreneurial activity: Geographic concentration of footwear production in the United States, 1940-1989,’ American Journal of Sociology 106, 424-62. Strange, W., Hejazi, W., Tang, J., 2006. The uncertain city: Competitive instability, skills, innovation, and the strategy of agglomeration, Journal of Urban Economics, 59, 331-351. Sveikauskas, L., 1979. Interurban differences in the innovative nature of production, Journal of Urban Economics 6, 216-227. Tiebout, C, 1956. A pure theory of local expenditures, Journal of Political Economy 64, 416-424. Vernon, R., 1960. Metropolis 1985, Harvard University Press, Cambridge, MA. Waldfogel, J., 2008. The median voter and the median consumer: Local private goods and population composition, Journal of Urban Economics 63, 567-582. Webster’s New World Dictionary, Second College Edition, 1970, G.C. Merriam & Co., Springfield, MA.

33

Appendix: Proofs.

Proof of Proposition 1:

If there are N workers in the city each of whom earns a wage of W, and if total

r l pro its equal , then the price of land will equal entrepreneu ia f

. The assumption of iso-elastic demand ensures that prices will equal

, where denotes the wage. Total labor earnings plus entrepreneurial profits in

the city will equal , so the price of land is times this amount. Total wealth

including renter income in city equals . If 0, total welfare for a

manufacturing worker in t ua the ci y eq ls:

(A1)

If 0, the city is a closed economy, wages can be normalized to one, and worker

welfare equals

. (A2)

Total consumption of each non-traded or non-traded good in the city equals

, where Y denotes the wealth of the relevant individual. The total

consumption of each good in the city is and consumption of all

domestically produced goods in the city equals . Demand for each traded

good outside the city is , so total exports in this sector equals . Total

labor de dman equals

(A3)

Setting th to l o esis equal ab r supply giv

= . (A4)

t ach non-traded goods entrepreneur equals The otal profits of e , (A5)

and the total profits of each traded good entrepreneur equal

34

. (A6)

From (A4), we have:

=

(A7)

Differentiating (A7) with respect to s yield

0. (A8)

Using the notation , (A8) can be rewritten

0. (A9)

Differentiating the equation with respect to elds yi

= 0. (A 0) 1

The numbers of workers per firm equals which is obviously declining with

when N n f o is fixed and indepe dent o . W rker welfare equals

, (A11)

Substituting in

= , (A12)

we obtain

. (A13)

The derivative of the logarith A 3) w th e ect to yields m of ( 1 i r sp

1 , (A14)

which is positive if

2 1 2 1 0 (A15)

The inequality (A15) always holds.

The derivative of the logarithm of (A13) with respect to yields:

, (A16)

35

which i f s positive if and only i 1. (A17)

(A17) always holds.

The cost of land equals: . It is increasing in since wages are

increasing in . The derivative of the logarithm of land costs with respect to is

,

which is positive if and only if .

(A18)

Proof of Proposition 2:

The spatial equilibrium requires that worker welfare equals the constant

reservation utility. This implies that or

. are constant with respect to changes in the

parameters and .

Standard notions of stability require that welfare be declining with N (taking into

account the impact that N has on wages). Differentiating the labor equilibrium equation

with res c o N yields: pe t t

= 0

(A19)

Since w a g pulation it is sufficient to ascertain that ages are alw ys declinin with po

. (A20)

is alway i i ogarithm of (A20) yields: s declining w th W. D fferentiating the l

1 . (A21)

For this l quantity to be a ways positive, it must be that

1 0. (A22)

Rearran oging and simplifying, the condition bec mes

2 1 1 0, (A23)

which must hold.

36

Differentiating the logarithm of (A23) with res ct to ie ds:

1

pe y l

1 0.

(A24)

This implies that:

The denominator is positive. The numerator, 2 1 1

1 1 , is obviously positive as long as 2 1 1 . If is

sufficiently close to one, it will be negative. Note that if 1, then

.

(A25)

must be constant with respect to changes in which ensures that wages will

fall as entrepreneurship rises.

Differentiating the logarithm of (A23) with respect to yields:

1

1 0, (A26)

which i plies m

0.

The cost of land equals a constant times , yet we know that

(A27)

is constant. This implies that

will be constant, and that will rise if and only

if

(A28)

rises. Th va v th ari h es o ith respect to yields: e deri ti e of e log thm of t is expr si n w

= , (A29)

which is always positive, so land values a l r wi re a ways ising th .

T e der ve o rh ivati f the loga ithm of with respect to is

. (A30)

(A31) is positive if

37

1

2 1 1 1 12 1 1 0

(A31)

(A31) ca nn be rewritte as

(A32)

which is always positive.

To find the effect of on N, it is enough to note that

st b c tant, so N will rise if mu e ons

(A33)

rises. The derivative of the logarithm of this expression with respect to is

(A34)

which is always positive.

To find the effect on N of , we first note that is rising with , so

if increases in lower W, it must be that N rises. As such we need only concern

ourselves with the cases where increases in increase W. Next differentiating

w th e to i r spect yields:

, (A35)

which must be positive if 0, which m t must increase. eans tha

Workers per firm is isomorphic to / yet we know

that must be constant when or changes.

Differen a n ti ti g the logarithm of with yields respect to

= , (A36)

which is positive if and only if 1 0. If 1 0, then

is increasing with and is decreasing with so

38

must rise as rises. If 1 0, then is decreasing with

and is increasing with so must rise s rises. a

Differentiating the logarithm f o with respec o yields:

1

t t

= . (A37)

If 1 0, then is decreasing with and is

decreasing with , so as rises must fall. However, if 1 0 it is

possible that may still be decreasing with if 1 1

1 2 1 , which would hold, for example for high enough levels

of . In which case, the comparative static would be reversed.

Finally, we turn to the profitability of the non-traded good entrepreneurs which

equals . The d tive of

eriva this with respect to is

1 2 1 , (A38)

or . (A39)

If 0 h r y i, t is exp ession is positive if and onl f

2 1 1 1 2 1 (A40)

which will always hold if is sufficiently close to one. Conversely if is distinctly

below one, then a large enough value of ensures that the inequality will fail.

T i e f r reneur profit with respect to is he derivat v o non-t aded entrep

2 , (A41)

which eq lua s:

.

This expression will be positive if is small and 1

1 1 , which will always hold if is sufficiently close to one. However, for

large values of the derivative must negative everywhere.

(A42)

39

Proof of Proposition 3:

The expected profit of each non-traded goods entrepreneur equals

l r, so the equi ib ium is that

. (A43)

For any u a , exogeno s v riable, Z we can write: , , , , (A44)

and tota f te th qu n t pect g lly di ferentia e a ti y with res to Z to et:

2 , (A45)

where we mean to signify the derivative of the

expression with respect to Z, not counting any indirect effects working through or .

Grouping terms together, this yields that equals:

(A46)

The expression 2 (A47)

represents the complete impact of on profits and since we have assumed that

, we are assuming that 1 and so the denominator is positive. As

such the sign of the derivative is the same as the sign of the numerator, or the sign of the

terms in parentheses since 0, which we refer to as M.

Case 1: Closed City

In the closed city case, wages are determined by the equality

40

= , (A48)

so the variable N, impacts wages but nothing else in the equation. Differentiating gives

< 0, (A49)

so entrepreneurship is rising with c size. ity

We will also investigate , and Q, and it helpful to note that in the closed city

case:

0

(A50) 0,

0.

In the case of , it is easy to see that the value of M is negative, since the wage effect is

positive and the direct effect on wages is also positive.

In the case of or Q, the value of M equals either or times

times minus

2 1

This can be rewritten as

2 1 (A51)

If is small and W<2 (which will occur if + 2 2 . then

this will hold. Alternatively, it will hold if is close enough to one. If is sufficiently

large, then wages will explode and all of the terms can be negative.

Case 2: Open City

In he city ase, wages are pinned down y the spatial equilit open c b brium

(A52)

41

42

This means that 0, so the number of entrepreneurs is again declining with , and

0, so the number of entrepreneurs is rising with Q. Also:

0, (A53)

so entrepreneurship is ith rising w .

The derivative 0, so the value of M

is

-

(A54)

This becomes:

*

1 + 1 1.

(A55)

As approaches one, the value of M approaches which is

unambiguously positive. Conversely if gets sufficiently high, then wages explode and

all of the terms are negative.